In absolute terms sure, but the token stream's confidence changes as it's coming out right? Consumer LLMs typically have a lot window dressing. My sense is this encourages the model to stay on-topic and it's mostly "high confidence" fluff. As it's spewing text/tokens back at you maybe when it starts hallucinating you'd expect a sudden dip in the confidence?
replyYou could color code the output token so you can see some abrupt changes
It seems kind of obvious, so I'm guessing people have tried this
Look up “dataloom”. People have been playing with this idea for a while. It doesn’t really help with spotting errors because they aren’t due to a single token (unless the answer is exactly one token) and often you need to reason across low probability tokens to eventually reach the right answer.
reply